ADXL345 on a magnet

Good morning, 

I'm using a ADXL345 for my project, its board is attached to a 30x2mm neodinium magnet with a non-conductive rigid glue. 

The sensor was brand new and working before attaching it to the magnet, now it's not. Facts: 

- by changing the orientation in space <x,y,z> of the sensor, the gravitational acceleration is not visible along the expected axis (not at all).

- the other digital registers I use (such as to adjust rate, sensitivity etc) are working as usual.

- the code I'm using is tested already on different "ADXL345"s, no problem

Do you have any idea why this setup is not working? My last theory regards possible damage from high shocks while attaching the magnet (and the sensor with it) on a metallic surface (even though 10'000g look unreachable, as per datasheet).

Thanks in advance

Parents
    •  Analog Employees 
    on Oct 15, 2019 3:21 AM over 1 year ago

    The magnetic field should not affect the accelerometer device operation since the sensor and package do not have any metals susceptible to the magnetic field, though we haven't specifically characterized for this.

    Based on the description you have on the failure sensor, does it recover when moved out from the magnetic field? Have you had a chance to test another sensor and see if it's reproducible? 

  • Thank you very much for answering!

    Unluckily it's not possible to remove the accelerometer from the magnet because the glue is permanent. 

    Some other details that may be useful: 

    - I started gluing 2 very same ADXL345 boards (same type of magnet, each sensor one of them) as I described earlier, the anomaly is equal in both right now. 

    - I tried to put the same magnet very close (almost attached) to a free accelerometer without gluing it, with this setup the sensor works as usual (as you say there is no visible changes in behaviour).

    thanks again, hope to hear from you soon

Reply
  • Thank you very much for answering!

    Unluckily it's not possible to remove the accelerometer from the magnet because the glue is permanent. 

    Some other details that may be useful: 

    - I started gluing 2 very same ADXL345 boards (same type of magnet, each sensor one of them) as I described earlier, the anomaly is equal in both right now. 

    - I tried to put the same magnet very close (almost attached) to a free accelerometer without gluing it, with this setup the sensor works as usual (as you say there is no visible changes in behaviour).

    thanks again, hope to hear from you soon

Children
    •  Analog Employees 
    on Oct 15, 2019 4:06 PM over 1 year ago in reply to Andrea.bevi

    It doesn't look like a sensor issue to me since you mentioned the unglued sensor which is exposed to the same magnetic has no performance change. Please keep in mind that the magnetic field may also have impacts on certain support functions such as power supplies, paths to your AFE circuity etc. I think you may also want to look into the possible change for those params.